Only by understanding how nervous systems process information and solve problems can we hope to deal with neurological problems at the cellular level. The principal goal of our research is to determine how neural circuits composed of populations of interneurons process information and control behavior. To reach this goal we have been developing a model system that serves as an intermediary between relatively simple invertebrate preparations and more complex vertebrate processing circuits. Invertebrate preparations have been invaluable to neurobiological studies because they allow precise cellular analysis of each neural element of the system. As a result, investigators have been able to describe these circuits in exquisite detail. However, it is not clear to what extent the principles that underlie these systems can be extrapolated to more complex vertebrate systems that employ large populations of interneurons to process sensory information and control behavioral movements. The escape system of the American cockroach serves as an intermediary between these two types of systems. It employs at least 14 individually identifiable giant interneurons to conduct sensory information anteriorly to processing centers. A population of at least 26 identifiable thoracic interneurons receive and process the information contained in the giant interneurons. Our work over the past grant period established a set of rules to relate connectivity and information transfer in the thoracic interneurons to readily apparent morphological cues. In the next grant period we will utilize intracellular recording and dye-injection techniques to extend these data to information transfer between the thoracic interneurons and the appropriate motor neurons. The motor consequences of the system will be described as will the neural activity that is responsible for those movements. The pattern of connections from specific thoracic interneurons to individual motor neurons will then be described. With these data we will be in an unique position to propose and test models for information processing within the population of parallel elements that make up this circuit. The principals that are uncovered from studying this approachable model system will aid in understanding basic integrative properties in the more complex systems found in mammals.